An additional comment on vapor trapping..

I've tried silica gel and various other standard dessicants.  The problem
does not seem to be fractionation during the trapping phase, but rather
during the heating phase.  Even at very high temperatures, it seems
virtually impossible to get all of the water back out -- and the resulting
fractionation is strong.

I'd be interested if anyone has done this successfully, as I agree with
Andy Campbell that it would be a convenient method for the field.  Indeed,
it would be quite easy to automate.

Eric Steig


~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
Eric J. Steig, Assistant Professor           tel: 215/573-5978 office
Department of Earth & Environmental Science  tel: 215/898-9890 lab
University of Pennsylvania                   FAX: 215/898-0964
251 Hayden Hall, 240 South 33rd Street       email: [log in to unmask]
Philadelphia, PA 19104-6316                  http://www.sas.upenn.edu/~esteig

On Thu, 31 Aug 2000, Andrew Campbell wrote:

> Dear All,
> I would like to sample atmospheric water vapor.  The studies that I have
> run across which have done this, have pumped air through a cold trap to
> freeze the water out.  To make this process easier to set up in the field,
> I would prefer to trap the water at ambient temperature.  Does anyone know
> of a way to trap the water on a desiccant that does not cause isotopic
> fractionation when the water is later released (presumably by heating) for
> hydrogen and oxygen isotopic analysis.
> Thanks,
> Andy
>
> **********************************************
> Dr. Andrew R. Campbell                       *
> Professor of Geology                         *
> Dept. of Earth and Environmental Science     *
> New Mexico Tech                              *
> Socorro NM 87801                             *
> Phone: 505-835-5327                          *
> Fax:   505-835-6436                          *
> e-mail: [log in to unmask]            *
> www.ees.nmt.edu                              *
> **********************************************
>